Genotoxicity of nanoparticles evaluated using the in vitro micronucleus assay, a review of recent data

Abstract

The in vitro micronucleus assay is a well-known and established component of the standard genotoxicity test battery. The growing use of nanomaterials around the world along with human exposure to them has increased the need for risk assessment with regard to safety, including potential genotoxicity. The in vitro micronucleus assay is one of the most used tests for evaluating the genotoxicity of nanomaterials. This review compiles studies since 2017 that performed assessments of micronucleus formation in vitro after cellular exposure to different nanomaterials. Genotoxicity of a broad range of nanomaterials including silver, cerium, zinc, gold, nickel, cadmium, titanium, carbon, and aluminum in different cell types were reviewed. While clear trends could be seen for some nanoparticle types like silver and cerium nanoparticles, others like gold nanoparticles showed mixed results. This review highlights the usefulness and effectiveness of the micronucleus assay for studying the genotoxicity of nanomaterials, in part, and is also careful to note that standard guidelines should be followed when conducting this assay in order to generate reliable and quality-driven data.

Introduction

Nanomaterials (NMs) have become an integral part of everyday life in recent decades, sometimes without much notice. Their prominence in technological advancements arises from their adjustable physicochemical properties, including melting point, wettability, electrical and thermal conductivity, catalytic activity, and light absorption and scattering. These characteristics provide enhanced performance compared to their bulk counterparts. As a result, NMs are widely used in manufacturing processes and consumer products, leading to an increased human exposure. Therefore, many genotoxicity studies on the safety of NMs have been conducted and provide useful data for risk assessment of NMs’ genotoxicity.^1–4 Depending on the type of NMs, different routes of exposure can ensue; some are in topical creams and are exposed directly through skin contact while others can be ingested or inhaled.

It has been recommended that mammalian cell micronucleus assay (MN) should be used for in vitro genotoxicity assessment of nanoparticles (NPs) to evaluate structural and numerical chromosome aberrations since mammalian cells are able to internalize NPs compared to bacterial assays.⁵ Thus, MN is a key component of both ICH S2(R1) and EFSA (2018) strategies to ensure the standardized evaluation in accordance with the OECD guideline. The MN assay is a sensitive and comprehensive test for the detection of chromosomal damage from either aneugens or clastogens in multiple cell types. The two most common types of MN tests, the cytokinesis block-micronucleus (CBMN) assay and the flow cytometry-based non-CBMN assay, are both currently applied in evaluations of NMs’ genotoxicity. CBMN requires that the exposed cell cultures are treated with Cytochalasin B (CytoB) to inhibit actin assembly which prevents further mitosis. After exposure and treatment with CytoB, cells are fixed and microscopically scored to identify binucleated and multinucleated cells.^6,7 The flow cytometry-based method uses fluorescent dyes to quickly detect a large number of MN as either binucleated or multinucleated cells, and to provide cytotoxicity information as well.^8,9

There are a substantial number of studies on the genotoxicity of NMs evaluated with the MN assay. We conducted a literature search in PubMed using the following keyword combinations, “nanomaterials/micronucleus” and “nanoparticles/micronucleus”. Due to the overwhelming number of search results, we have chosen manuscripts published within the last eight years (2017–2024) and studies conducted using the well-established standard in vitro MN assay with commonly used NMs. For earlier studies, readers are encouraged to pursue previous reviews on MN assay evaluation of NMs.^2,10,11

Compilation of studies

Silver nanoparticles (AgNPs) are one of the most utilized NPs that are applied in various consumer products including medicine, cosmetics, textiles, and food products due to their strong antimicrobial activity. Humans can be directly exposed to AgNPs through multiple routes, including inhalation, dermal contact, and oral ingestion. The exposure routes, along with physicochemical properties such as particle size, surface charge, and coatings, plays a crucial role in AgNP uptake by different cell types and their interaction with biomolecules. Consequently, the genotoxicity of AgNPs have been extensively studied in various cell models under distinct conditions. This review includes in vitro studies assessing the effects of AgNPs on MN formation (Table 1). Most of them reported significant increases in MN formation at one or multiple concentrations of AgNPs, regardless of coating, across various cell models. In a recent study, BEAS-2B cells were treated with uncoated AgNPs at concentrations ranging from 1 to 25 µg/mL. The MN frequency was increased in the presence of cytochalasin B after 48 h of exposure at most concentrations.¹² The cellular uptake of AgNPs was also confirmed via Transmission electron microscopy (TEM).¹³ It was also shown that the uncoated AgNPs could increase MN formation by activating the GADD45a gene and then affecting the DNA damage response pathway in human liver carcinoma HepG2 cells and human adenocarcinoma alveolar basal epithelial A549 cells.¹⁴ Other cells, including CHO-K1, JURKAT, and TPH-1, have shown positive results in in vitro MN following exposure to uncoated AgNPs at varying concentrations for 24 or 48 h.^13,15–19 After exposure of 5 nm PVP-coated AgNPs at concentrations ranging from 1 to 1.5 µg/mL to TK6 cells, an elevated production of hydroxyl radicals and ROS was observed. Additionally, a concentration-dependent increase in cytotoxicity and MN formation was detected 28 h after the treatment.²⁰ A different study also using the PVP-coated AgNPs at concentrations ranging from 0.5 to 75 µg/mL demonstrated increased frequencies of micronuclei in AgNP-treated cell populations in human hTERT-immortalized retinal pigmented epithelial (RPE-1) cells. Moreover, MN induction with different exposure times resulted in a different nuclear phenotype and further cell division defects.^14,21 Additionally, PVP-coated AgNPs were found to enhance MN formation in both human lymphoblastoid TK6 cells and mouse lymphoma L5178Y cells.²² Another mouse fibroblast cell line Balb/c 3T3 A31-1-1 cells were treated with PVP-coated AgNPs at 10.6 μg/mL for 24 h, resulting in increased MN formation and neoplastic transformation and transformation frequency.²³ In a 2021 study using a 3D model revealed that treatment of PVP-coated AgNPs at 0.5, 1, and 5 μg/mL for 24 h resulted in a significant increase in MN frequency. With another common coating, citrate, citrate-coated AgNPs, at concentrations of 10 and 40 μg/mL were found to induce a significant increase in MN formation in keratinocyte HaCaT cells, although no significant increase in neoplasmatic bridges or nuclear buds was observed.²⁴ L-methionine can also be used as the stabilization agent to coat AgNPs. A study showed that exposure to L-methionine coated AgNPs significantly induced MN formation; there was, however, no significant reduction in catalase activity or increase in malondialdehyde (MDA) protein levels, suggesting that the L-methionine-coated AgNPs did not promote oxidative stress under the experimental conditions.¹³ However, some studies also presented conflicting evidence regarding the genotoxicity of AgNPs. Specifically, neither small (5 nm) nor large (50 nm) citrate-coated AgNPs, at concentrations from 1 to 20 μg/mL for 48 h, were able to increase MN formation, although they induced DNA damage measured via the mini-gel Comet assay.²⁵ In a study in 2020, no elevation in MN frequency over the negative control were observed following exposure to 35 nm or 50 nm PVP-coated AgNPs at concentrations from 0.012 to 12 μg/mL in human blood peripheral lymphocytes while 50 nm PVP-coated AgNPs significantly decreased the replication index and significantly increased cytostasis, apoptosis, necrosis, and the frequencies of nuclear buds and nucleoplasmic bridges.²⁶

Table 1.

Summary of genotoxicity of nanomaterials evaluated using in vitro micronucleus assay.

Nanomaterial	Coating	Dose	Characterization included or referenced	Cell type	NP uptake*	Media aggregation included	Result	Assay type	OECD guidelines fully followed
AgNPs	L-methionine13 PVP20–23,26,66 Citrate24,25 Uncoated12,14–19	0.1–10 pM13 12.5–200 µg/ml14 0.012–12 µg/ml26 1–30 µg/ml22 0.17–10.6 µg/ml23 0.5–75 µg/ml21 1–25 µg/mL12 1.5 µg/ml20 0.2–10 µg/ml66 10–40 µg/ml24 1–20 µg/ml25	Yes12–26,66	Human blood peripheral lymphocytes26 HepG2 and A54914,66 L5178Y, TK6 and CHO20,22 Balb/c 3T3A31-1-123 HaCaT cells24 BEAS-2B12 RPE-121 HBEC3-kt25	Preferential adsorption13 Cell viability14,15,17–20,22–26,66 TEM12,16,21,66 ICP-MS21	Yes12–17,20,23,24,26,66 No18,19,21,22,25	Increased MN formation12–24,66 No effect on MN formation25,26	Cytokinesis Block MN assay (CB MN)12,14,15,18,19,21,23,24,26,66 Flow cytometry assay16,17,20,22,25	21,23,24
CeO2 NPs	Uncoated27–33 PVP62	10–200 µg/ml27 10–100 µg/ml29,30 100–300 µg/ml31 20–100 µg/cm233 5–40 µg/ml62 0.78–50 ppm28 6–18 µg/ml32	Yes27,31,33,62 No29,30	Human peripheral blood culture28 A54933 Peripheral blood lymphocytes29,32 PC-3 cells31 HL-6030 IMR3227 FE162	Cell viability27,31,33,62 Flow cytometry29	Yes33,62 No27,29–31	Increased MN formation27–32 No effect on MN formation33,62	CBMN27–33 Flow cytometry62	32,33
ZnO NPs	Humic acid/HA-like polycondensate34 Gadolinium41 Uncoated35–40	0.5–20 µg/ml34 10, 20 µg/ml41 0.5–10 µg/ml36 1–20 µg/ml37 109–172 uM38 1–2000 ppm39 12–123 uM40 12–307 µmol/L35	Yes34–41	Human lymphocytes34 SKLC-6 cells41 HaCaT cells36 V-79 cells37 CHL/IU cells38 Human peripheral blood lymphocytes39 MDCK cells40 Caco-2 and LT9735	Cytotoxicity34 TEM and ICP-MS41 TEM and flow cytometry36,37 TEM and cell viability38 Ion release cytotoxicity39 ICP-MS and cell viability35	Yes34–37,40,41 No38,39	Increased MN formation35–40 No effect on MN formation34,35,41	CBMN34,38–41 Flow cytometry35–37	40
AuNPs	L-methionine13 Citrate25,43,47 Carboxylate44 Ammonium44 PEG44,47 4ATP45 IP6-Jacalin46	0.1–10 pM13 6.2–50 µg/ml43 1–256 µg/ml44 1.2 nM45 10 uM, 15 uM46 0.5–5 nM47 1–20 µg/ml25	Yes13,25,43–45,47 No46	Human peripheral blood cells13 CHO cells43 BEAS-2B cells44 NIH-3T3 cells45 HCT-1546 NCM 46046 HBEC25 Caco-2 and HaCaT47	Preferential adsorption13 Microscopy/cell viability43,46 Hyperspectral Imaging44 SR-micro FTIR45 Cell viability25 Darkfield images47	Yes13,25,43–45,47 No46	Increased MN formation44,46,47 No effect on MN13,25,43–45	CBMN13,43,45,47 Flow cytometry25,44,46	43
NiO NPs	Pure NiO NPs48–50	1–10 µg/ml48 15–500 µg/ml49 5–25 µg/ml50	Yes48–50,48,49,67	BEAS-2B48 V-7949 THP-1 and HBEC50	TEM and ICP-MS48 TEM49 ICP-MS50	Yes48,49 No50	Increased MN formation48–50 No effect on MN formation50	CBMN48,49 Flow cytometry50	48
Cd NM CdO NPs Quantum Dot NPs	PEG51 Glutathione52 Mercaptosuccinic acid53	1–10 µg/ml51 0.5–10 µg/ml52 50–200 nM53	Yes51,52	TK651 CHO52 Mollusk hemocytes53	Ion release/cell viability51 ICP-MS53	Yes51 No53	Increased MN formation51,52 No effect on MN formation53	CBMN52,53 Flow cytometry51	51
TiO2 NPs	Uncoated22,55–57	10–800 µg/ml22 1–25 µg/ml56 3–75 µg/cm355 0.14–14 µg/ml57	Yes22,55–57	TK622,55 HUVEC56 HT29-MTX-E12 and Caco-257	Cell viability22 DNA damage55,56 TEM57	Yes22,55,57 No56	Increased MN formation56,57 No effect on MN formation22,55,57	CBMN22,55–57	55
Carbon NM Nanotubes Nanofibers	Uncoated58–61	5–50 µg/ml59 1–100 µg/ml60 1.5–50 µg/cm261 25–125 µg/cm258	Yes58–61	BEAS-2B59 A54958,60,61	Cell viability58,59 Flow cytometry60 TEM61	Yes58–61	Increased MN formation58–61	Flow cytometry59–61 CBMN58	61
Al2O3 NPs	Uncoated39,62	1–250 ppm39 5–40 µg/ml62	Yes39,62	Human peripheral blood lymphocytes39 FE162	TEM and cell viability62 Not assessed/not listed39	No39,62	Increased MN formation62 No effect on MN formation39	CBMN39 Flow cytometry62	N/A

Abbreviation: NM – nanomaterial; NP – nanoparticle; AgNPs – silver nanoparticles; TiO₂ NP – titanium dioxide nanoparticles; ZnO NPs – zinc oxide nanoparticles; AuNPs – gold nanoparticles; CeO₂ NPs – cerium oxide nanoparticles; CdO NPs – cadmium nanoparticles; NiO NPs – nickel nanoparticles; Al₂O₃ NPs – aluminum oxide nanoparticles.

Cerium oxide nanoparticles (CeO₂ NPs) have been widely used because of their unique surface chemistry, high stability, and biocompatibility. It is mostly utilized as a polishing material, an additive in glass and ceramics, fuel cell materials, agricultural products, and a catalyst/additive in automotive fuels. Thus, CeO₂ NPs can be released into the ambient air and other natural ecosystems. Increased MN formation following CeO₂ NP exposure has been observed in several in vitro MN studies using different cell models. One study reported a concentration-dependent increase in MN formation and a decrease in cell proliferation. At 100 and 200 µg/mL, MN frequency per 1000 binucleated cells increased significantly with the treatment. Additionally, the cell proliferation index was significantly reduced at 50, 100, and 200 µg/mL compared to the control. These results suggest that CeO₂ NPs induce genotoxicity and impair cell proliferation in a concentration-dependent manner.²⁷ Exposure to CeO₂ NPs at concentrations as low as 0.78–50 ppm for 72 h induced MN formation in human peripheral blood cultures.²⁸ CeO₂ NPs were examined as a possible protective agent due to their radioprotective attributes and radical scavenging properties.²⁹ However, it was found that CeO₂ NPs themselves can significantly increase MN formation in human lymphocytes at a concentration of 100 µg/mL, human leukemia HL-60 cells at concentrations of 10, 25, 50, or 100 µg/mL,³⁰ and human prostate cancer cell line PC-3.³¹ In another study evaluating the genotoxicity of CeO₂ NPs, MN formation increased in a concentration-dependent manner at concentrations of 6, 12, and 18 µg/mL following short-term exposure (3–24 h) to human peripheral blood lymphocytes.³² In this study, relatively lower concentrations and shorter exposure duration were employed, highlighting the sensitivity of the assay in detecting genotoxic effects of CeO₂ NPs. In contrast, several other studies reported conflicting results, with no observed increase in MN formation following exposure of uncoated or coated CeO₂ NPs to several different cell lines. The authors attributed these negative findings to inconsistencies in the particle size of the test articles.³³

Zinc oxide nanoparticles (ZnO NPs) are widely used in manufacturing processes for electronics, as additives in catalysis, and textiles as well as cosmetics, particularly in sunscreens due to their photocatalytic properties. The MN results from the majority of the literatures suggested that ZnO NPs were genotoxic. A study examining the effects of ZnO NPs in combination with humic acids (HAs) on human lymphocytes found a slight increase in MN frequency at the highest concentration of ZnO NPs (20 μg/mL) while the combination of ZnO NPs with HAs did not result in significant increase in MN induction compared to the control group, suggesting no genotoxicity associated with the combination of ZnO NPs and HAs.³⁴ The determination of cell models impacts the observed genetic effects of ZnO NPs at comparable experimental conditions. One study reported a significant increase in MN formation in the human colon adenoma cell line LT97 at 307 µmol/L ZnO NPs (< 50 nm), whereas no such effect was observed in the human colorectal adenocarcinoma cell line Caco-2.³⁵ An investigation into ZnO NP treatment in HaCaT cells revealed a significant increase in MN formation. Notably, this increase was both concentration- and time-dependent after 3 and 6 h of treatment at concentrations 2.5, 5, and 10 μg/mL. The genotoxic potential of ZnO NPs was further demonstrated by elevated intracellular ROS levels and a decrease in DNA damage response.³⁶ Consistent with these findings, another study on ZnO NPs in hamster lung V79 cells reported an increase in MNs in a concentration-dependent manner (1–20 μg/mL) after treatment along with an increase in ROS formation as well as cell cycle arrest, indicating the clastogenic and aneugenic effects of ZnO NPs.³⁷ A study in Chinese hamster lung CHL/IU cells demonstrated that ZnO NPs caused increased MN formation in the presence or absence of rat liver S9 mixture. This increase in MNs could be completely blocked by introducing the antioxidant tempol, further implicating ROS generation as the genotoxic mechanism.³⁸ ZnO NPs (12.5–250 ppm) were also shown to significantly increase the production of MNs in human blood lymphocytes. However, at 500, 1000, and 2000 ppm, ZnO NPs induced overt cell death, preventing MN frequency assessment at these concentrations.³⁹ A study comparing the genotoxicity of ZnO NPs and ZnCl₂ showed that only ZnO NPs increased MN formation in Madin-Darby canine kidney (MDCK) cells. Although no direct elevation in ROS was observed, both glutathione transferase and catalase activities were significantly reduced,⁴⁰ suggesting that impaired intracellular defense mechanisms against oxidative damage directly contribute to MN formation. Interestingly, the genotoxicity and cytotoxicity mechanisms of ZnO NPs have been linked to their photocatalytic properties under UV light. One study found that although gadolinium-coated ZnO NPs alone did not significantly increase MN formation, their combination with X-irradiation resulted in an additive effect, leading to greater MN formation compared to X-ray exposure alone.⁴¹

Gold nanoparticles (AuNPs) have been widely used in biomedical science, including cancer cells diagnosis, cancer therapy, and HIV treatment.⁴² To enhance their stability, functionality, and biocompatibility, various coatings have been applied, such as polyethylene glycol (PEG), biomolecule, and small molecule coatings like citrate. Given their expanding applications, an increasing number of toxicological evaluations have been conducted on AuNPs across various test systems. The MN data showed that MN test results for AuNPs were less consistent compared to those for AgNPs, CeO₂ NPs, or ZnO NPs, with positive and negative findings being evenly divided. L-methionine-capped AuNPs showed no effect on MN formation and did not impact catalase activity or malondialdehyde levels in human peripheral blood cells but did demonstrate the ability to suppress cell proliferation.¹³ Treatment of the normal human bronchial epithelial cells HBEC with 1, 10, and 20 µg/mL of citrate coated AuNPs with different diameters did not induce MN formation and DNA damage, nor reduced cell viability.²⁵ Similarly, in another study using CHO cells, exposure to AuNPs at 6.2, 12.5, 25, and 50 µg/mL for 20 h did not increase MN formation.⁴³ However, several studies on AuNPs showed conflicting results on MN induction. In a study on AuNPs in lung epithelial BEAS-2B cells, different coatings and sizes produced varying effects on MN formation. Small (5 nm) ammonium coated AuNPs significantly increased MN frequency at concentrations starting from 1 µg/mL, while larger (20 nm) ammonium coated AuNPs did not significantly induce MN formation. Carboxylated AuNPs increased MN frequency only at the lowest tested dose (5 µg/mL), while PEGylated AuNPs induced MN formation at doses exceeding 1 µg/mL. This study suggests that size, coating, and cellular uptake play critical role in the genotoxicity of AuNPs.⁴⁴ In a study on 4-aminothiophenol-coated AuNPs in fibroblastic NIH-3T3 cells, no increase on the frequency of micronucleated cells was found following a short-term exposure for 15 or 30 min. Moreover, 4-aminothiophenol-coated AuNPs exhibited protective effects, as co-exposure to NPs and ultrasound for 30 min resulted in a reduced number of micronucleated cells compared to ultrasound exposure alone.⁴⁵ With another type of coating, IP6-Jacalin-coated AuNPs at concentrations of 10 and 15 μM were found to increase MN, impact the cell cycle, and induce ROS production as the mechanism of AuNPs’ genotoxicity.⁴⁶ In a recent study, Caco-2 and HaCaT cells were treated with various types of AuNPs, including citrate-stabilized AuNPs, PEG-liganded carboxyl AuNPs, PEG-liganded amine AuNPs, and PEG-liganded hydroxyl AuNPs, at concentrations of 0.5, 1, 2, and 5 nM – relatively lower than doses used in previous studies discussed here – for 2 h, resulting in a concentration-dependent increase in micronuclei and nucleoplasmic bridges.⁴⁷

Nickel oxide nanoparticles (NiO NPs) are a significant class of transition metal oxides with a cubic lattice structure, making them highly versatile for various applications. These NPs have unique properties, including optical, thermal, electrical, physicochemical characteristics. Due to their exceptional properties, NiO NPs have been widely used in solar and lithium batteries, photoelectron devices, ion storage materials, sensors, magnetic and thermoelectric materials, catalysts, electrochromic products. Additionally, NiO NPs exhibit biomedical potential, including anticancer, cytotoxic, antibacterial, and non-enzymatic glucose-sensing properties. As one of the most used nanoparticles, it is essential to review its genetic effects. Unlike the mixed test results observed with AuNPs, these studies on NiO NPs consistently concluded that the NPs were genotoxic and contribute to increased MN formation. NiO NPs at concentrations of 5 and 10 µg/mL for 48 h increased MNs, intracellular ROS, and DNA damage in BEAS-2B cells, while also elevating intracellular calcium.⁴⁸ Another study exposed V79 cells to various concentrations of NiO NPs at 15, 31, 62, 125, 250, 500, 1000, and 2000 μg/mL; and significant differences in the frequency of binucleated micronucleated cells (BNMN) were observed after the treatment at concentrations of 250 and 500 μg/mL for either 4 or 24 h. Furthermore, NiO NPs induced an increase in MNs in V79 cells, which was likely associated with DNA damage caused by ROS production.⁴⁹ A study evaluated both primary and inflammation-driven secondary genotoxicity of NiO NPs. It was found that NiO NPs induced MN formation in THP-1 macrophages exposed to 25 μg/mL for 48 h on inserts, but not in human bronchial epithelial cells (HBEC-3kt) cultured in the lower compartment of the insert to assess indirect effects. However, DNA strand breaks were observed after indirect exposure to NiO NPs in HBECs.⁵⁰

Cadmium (Cd) nanomaterials (NMs) are widely used nanoparticles (NPs) found in paint pigments and electronics. They are also the starting material for the manufacture of Cd quantum dots which are being explored for their diagnostic imaging and therapeutic relevance. Cadmium oxide nanoparticles (CdO NPs) can be applied for optical coatings, photovoltaic cells, phototransistors, IR reflectors, transparent electrodes, and gas sensors. In a recent publication, PEGylated CdO NPs significantly increased MN formation in TK6 cells at concentrations of 5, 7.5, and 10 μg/mL for 4 h. Furthermore, the data showed that CdO NPs caused genotoxicity directly rather than through the released Cd⁺.⁵¹ A study evaluated genotoxicity of glutathione-capped Cd quantum dots in CHO cells using the MN assay. The results showed that the quantum dots significantly increased the frequency of MNs at concentrations ranging from as low as 0.5 μg/mL to 10 μg/mL, compared to the vehicle control level.⁵² Another study investigating the effects of cadmium telluride (CdTe) quantum dots on mollusk hemocytes found no significant MN induction at concentrations of 50, 100, and 200 nM. However, the mollusk Biomphalaria glabrata used as the model in this study may behave differently compared to other cell lines in in vitro assays. The toxicity of quantum dot suspensions to B. glabrata may involve various mechanisms, and it may not be directly linked to the presence of Cd in tissues.⁵³

Titanium dioxide NPs (TiO₂ NPs) have broad industrial applications due to their low cost, excellent chemical stability, high refractive index, and strong oxidation properties. They are widely used in the chemical, biomedical, and solar energy fields, as well as in products such as sunscreens, lacquers, and paints. Studies on the genotoxicity of TiO₂ NPs have yielded inconsistent results. According to Charles et al., from 2010 to 2016, approximately 35% of studies using the MN assay reported positive findings.⁵⁴ This paper reviewed four studies on the effects of TiO₂ NPs on MN formation, revealing mixed results. TiO₂ NPs at concentrations of 10–800 μg/mL exhibited cytotoxicity but did not significantly induce micronuclei, whereas 0.75 Gy X-ray exposure led to an expected > 11-fold increase in MN formation in TK6 cells.²² A different in vitro MN study on TiO₂ NPs (3, 15 and 75 μg/cm²) for 24 h reported no significant changes in MN formation in peripheral lymphocytes and TK6 cells, despite observing DNA damage at the highest administered dose.⁵⁵ Additionally, no cytotoxicity was observed at any NP concentration (3, 15 and 75 μg/cm²) tested. Although the highest concentration did not reach the 2 mg/mL limit recommended by OECD guidelines, the absence of cytotoxicity and genotoxicity across tested concentrations supports a negative result. However, in a study utilizing different sizes of TiO₂ NPs at concentrations of 1–25 μg/mL led to an increase in MN formation, which was associated with elevated intracellular ROS levels, reduced GSH levels, and increased NRF2 protein expression in umbilical vein endothelial HUVEC cells.⁵⁶ In a recent study, three different TiO₂ NPs increased MN frequency in HT29-MTX-E12 cells, with the effect being more pronounced after in vitro digestion, suggesting potential genotoxicity. In contrast, no significant MN increase was observed in Caco-2 cells, except for one type of TiO₂ NPs, undigested NM-105 at 14 μg/mL, highlighting cell-specific responses to these NPs.⁵⁷

Carbon nanomaterials (carbon NMs), including graphene, carbon nanotubes, crystalline diamond, and diamond-like carbon, exhibit exceptional electrochemical properties, leading to their widespread applications. In vitro MN studies on different carbon NMs were reviewed here. Multiwalled carbon nanotubes (MWCNTs), which show promise for use in drug delivery, photothermal, and imaging applications, were tested for MN formation in BEAS-2B and A549 cells. The selected MWCNTs at concentrations of 100 and 125 µg/cm² induced MN formation in A549 cells, however, negative effects were observed in the BEAS-2B cells via the CBMN assay.⁵⁸ While another study demonstrated positive MN induction in BEAS-2B cells at 20 and 50 μg/mL using an MN assay, though without a clear concentration-response relationship.⁵⁹ Genotoxicity of graphite nanofibers that are used as electrodes in electrochemical cells and biomedical scaffolding was evaluated and the effect of graphite nanofibers on MN formation in A549 cells was measured; and they significantly increased MN formation; genotoxicity was mediated through production of ROS along with disruption of autophagy pathways.⁶⁰ In a study examining multiple different carbon NMs, exposure to cellulose nanofibrils (6 and 12.5 µg/cm²), cellulose nanocrystals, and multi-walled carbon nanotubes-402 did not significantly alter micronucleated binucleated cell (MNBNC) frequency in A549 cells. However, cellulose microfibrils (1.5 and 50 µg/cm²) and multi-walled carbon nanotubes-401 (all tested concentrations) significantly increased MNBNC frequency, with multi-walled carbon nanotubes-401 having the strongest effect.⁶¹

Lastly, two study on aluminum oxide NPs (Al₂O₃ NPs) were reviewed here. Al₂O₃ NPs are used frequently in the production of electronics as well as in fillers used in cosmetics. In human peripheral blood lymphocytes, Al₂O₃ NPs did not increase MN formation, nor did they create additional chromosome aberrations. The authors attributed discrepancies between their results and previous in vivo studies on Al₂O₃ NPs to the size of the nanoparticles used.³⁹ Uncoated Al₂O₃ nanoparticles induced a 2.1-fold increase in MN formation at 20 and 40 µg/mL, whereas Al₂O₃ microparticles led to 1.2- and 1.8-fold increases at the same concentrations.⁶² The differences in genotoxicity suggest mechanistic variations in the response to Al₂O₃ NPs, potentially influenced by their particle sizes.

Quality assessment

Every study included here, with one exception, performed a characterization of the NMs used or cited a previous manuscript that used the exact same materials. Characterization of NPs relied heavily on three main pieces of instrumentation, the transmission electron microscope, a UV-visible spectrometer, and a Zetasizer Nano that measures zeta potential and particle size. If the characteristics of the NPs are not listed within the Materials and Methods section of the manuscript, they are often displayed in a table containing categories like coating, size distribution, hydrodynamic diameter, and Zeta potential, among others. While NP characterizations were included in all but one study, a large discrepancy exists when examining NP aggregation in cell culture media used. Naturally, with a broad range of cell types used in each of the reviewed in vitro MN studies, many different types of media were utilized. However, only approximately two-thirds of the studies included information on NP aggregation within the culture media alone. Additionally, most studies provided data regarding true NP uptake into cells. Table 1 shows those references broken down by study.

The OECD Test Guideline 487 contains specific recommendations for performing the in vitro MN assay to obtain interpretable and reliable results. Currently the in vitro MN assay is designed to measure non-nanomaterials to assess the cytotoxicity and genotoxicity of NMs. Due to the different natures and types of NMs, they can interfere with the performance of in vitro MN assays. There are specific factors that can affect the cytotoxicity and genotoxicity assessment of NMs when measured using the in vitro MN assay, such as the choice of cell line, NP coating, diameters of NPs, fluorescence of NMs, the use of CytoB, fetal bovine serum in cell treatment medium, and different measurement methodologies for micronuclei. There are several well-documented examples of the interference of NMs in assay performance. Two prominent examples involve TiO₂ NPs. It was shown that TiO₂ NPs impact the CBMN specifically due to a critical component of the assay itself, CytoB, which prevents uptake of TiO₂ NPs.⁶³ TiO₂ NPs were also shown to affect the cell counter and flow cytometer measurements in a flow cytometry based MN.⁶⁴ Hence, choice of cell type, selection of methodology, and modification of existing methods should be carefully considered before genotoxicity evaluation of NMs using an in vitro MN assay.

In addition, there should be a cytotoxicity measurement of test agent included in the published results since the assay itself is dependent on mitotic cell division. A recommendation to use a minimum of three test concentrations is listed, along with relevant vehicle and positive controls which have also be subjected to cytotoxicity tests. The highest test concentration should be the lowest relevant value of the following: 10 mM, 2 mg/mL, or 2 µg/µL. In this review all studies that reported a negative result in the MN assay did not use the relevant highest test concentration, therefore, these data did not meet the criterion for claiming that the NMs tested were not genotoxic in the MN assay. Though it should be noted that OECD-487 indicates that specific adaptations to the test guideline are needed for NMs, though they are not described in detail.

Conclusions

The in vitro MN assay is an excellent tool for genotoxicity assessment of test substances due to its sensitivity and specificity for detecting oxidative DNA damage induced by NMs. In this review, the results of MN assays from several different types of NMs across multiple and varied cell lines were compiled from the past eight years (2017–2024). As expected, results varied depending on NP type and cell culture environment. A consensus opinion can be formed for some of the NMs that have been more frequently examined. Specifically, AgNPs, CeO₂ NPs, and ZnO NPs induced MN formation in the majority (>70%) of the studies reviewed. Conversely, NMs, such as AuNPs, showed highly mixed results. In our opinion, for a study on the genotoxicity of a NM to be successful, it should include the following: (1) characterization of the NM or a recent reference to characterization of the exact material tested; (2) confirmation of NM uptake into the cells, either through electron microscopy or applicable imaging techniques; and (3) evaluation of NM aggregation in the cell culture medium. These parameters, in addition to adherence to the OCED guidelines of the assay, will enhance the quality of genotoxicity risk assessments involving NMs. For example, fewer than 30% of the studies cited in this review fully complied with the OECD test guidelines. While all the studies included successfully measured MN formation in an appropriate context, the most frequent failure to adhere to guidelines was the lack of either Relative Increase in Cell Count (for flow cytometry) or Proliferative Index/Relative Increase (CBMN). The studies reviewed here often reported cytotoxicity data; however, OECD-487 explicitly states that this information is useful “but should not be used in place of CBPI or RI”.⁵ While the in vitro MN assay is an effective component of the genotoxicity testing battery, it should not be considered sufficient on its own for determining genotoxicity. Although this method is relatively fast and cost-effective, other genotoxicity assays that can determine different genotoxicity endpoints such as point mutations should also be applied for risk assessment of NMs. Clearly, the use of this established methodology as part of a collective genotoxicity test series, following OECD-487 guidelines, and implementing the recommendations for NMs testing² will improve the quality and reliability of published data regarding potential human health risks associated with NMs.

Despite the widespread application of the in vitro MN assay for detecting chromosomal damage induced by NMs, several critical gaps and challenges remain in the test. One major limitation is lack of standardization across laboratories in terms of NMs characterization, exposure conditions, and assay implementation. In this review, we compiled and compared variables, such as nanoparticle concentrations, surface coatings, cell types, and exposure durations, used in the literatures. The findings from these studies were inconsistencies, which makes it difficult to perform cross-study comparisons and establish clear and consistent relationships between specific nanomaterials and cellular responses. Therefore, consideration should be given in future studies on dispersion methods, dosimetry, and key physicochemical parameters such as surface charge, agglomeration, and protein corona formation, as well as defined concentration ranges and exposure durations for particular cell models. For example, proper dispersion techniques should be used to ensure that NMs are stable and uniformly suspended, with at least one test concentration thoroughly characterized. Evaluation of cellular uptake is also important because it can determine whether the test NMs failed to induce a genotoxic response or simply did not reach the target cells when a test is negative.

Although the MN assay effectively detects chromosomal damage, it does not capture other important genotoxic endpoints such as mutations. Therefore, future investigations should consider combining the MN assay with one or more of assays for detecting mutagenicity of NMs to enable a more comprehensive evaluation of the genetic effects of NMs. Since the Ames test, the mutation test currently used for regulatory purposes, is not suitable for evaluating NMs due to issues with cellular uptake, emerging techniques, such as Duplex Sequencing and single-cell sequencing, should be utilized to enhance both sensitivity and specificity in detecting rare mutational events and provide mechanisms of action. Furthermore, artificial intelligence and machine learning approaches are emerging as powerful tools for interpreting the complex datasets generated by high-content genotoxicity assays. These technologies together with the NM assay should be considered in future genotoxicity studies.⁶⁵ Finally, selecting an appropriate and sensitive cell system is essential. The optimal model should origin from human, p53-competent, and exhibit a stable, low background frequency of micronuclei. As discussed in this review, several studies have already explored the use of 3D cell culture to improve the human relevance of in vitro findings. In particular, 3D models, organoids, and co-culture systems are increasingly recognized for their ability to better replicate in vivo tissue architecture and intercellular interactions. Incorporating these advanced models into NM genotoxicity testing could significantly enhance the predictive value of in vitro assays and help reduce reliance on animal testing.